Call for Research Interns �Jan/Feb 2026
About the Wearable and Interactive Technology Lab
The Wearable and Interactive Technology Lab in the School of Electrical Engineering at KAIST focuses on the design, development, and evaluation of wearable, physical, and tangible interactive computing systems. Bringing together design perspectives, computer science skills, and psychological methods, the WIT lab conceives, creates, and studies the next generation of human-computer interfaces.
SonarID: Sensing fingers around a smartwatch
FingerText: Typing on fingernails for AR
Gestures passwords for device lock
Why join WIT Lab?
Conduct research on emerging technology
Work with a graduate student mentor
Work towards a research paper
How to apply?
Egocentric Full-Body Motion Capture via Smart Glasses
Mentor: Hyunyoung Han (hyhan@kaist.ac.kr, Website)
Required skills: 3D modeling
Background: This research explores novel approaches to egocentric motion capture using commercially available smart eyewear. We will investigate optical and computational methods to expand sensing capabilities within the constraints of head-mounted devices, enabling full-body motion tracking without requiring instrumented environments or additional wearable sensors [1, 2]. The system addresses key challenges in self-occluded body tracking [3] through customizable hardware configurations and specialized computer vision techniques.
Expected Outcomes: Prototype for egocentric motion capture system, research paper
References
[1] Kang, T., Lee, K., Zhang, J., & Lee, Y. (2023, December). Ego3dpose: Capturing 3d cues from binocular egocentric views. In SIGGRAPH Asia 2023 Conference Papers (pp. 1-10).
[2] Dai, P., Zhang, Y., Liu, T., Fan, Z., Du, T., Su, Z., ... & Li, Z. (2024). Hmd-poser: On-device real-time human motion tracking from scalable sparse observations. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 874-884).
[3] Zhang, S., Ma, Q., Zhang, Y., Aliakbarian, S., Cosker, D., & Tang, S. (2023). Probabilistic human mesh recovery in 3d scenes from egocentric views. In Proceedings of the IEEE/CVF International Conference on Computer Vision (pp. 7989-8000).
Self-occlusion problem
Previous approaches to egocentric motion capture
Reading Webtoons in XR
Mentor: Ammar Al-Taie (ammar@kaist.ac.kr)
Required skills: Unity Development
Background: Webtoons are popular because they use familiar smartphone interaction techniques—scrolling, swiping, and tapping—to navigate panels. However, extended reality (XR) glasses are expected to replace smartphones as they become smaller and lighter. In this project, we will research how webtoons can be presented and navigated in XR, leveraging 3D object rendering, spatial affordances, and a range of sensors, including eye and face tracking. We will investigate how we can use XR to read webtoons while walking or travelling in public transport.
Expected Outcomes: Prototype XR webtoon concept and an experiment testing the concept.
Converting Webtoons to XR
Expanding the Design Space of Running Interfaces
Mentor: Ammar Al-Taie (ammar@kaist.ac.kr)
Required skills: Unity Development or Electronics Prototyping and passionate about running 🏃
Background: Running is the most popular physical activity worldwide. However, current devices, such as smartwatches or earbuds, are constrained to small screens or simplified audio and vibration feedback. In this internship, we will investigate the types of tasks runners do while running, and research how we can improve these tasks using new devices. We will develop a prototype device and then conduct an experiment to test it with real runners.
Expected Outcomes: Prototype running device, and an experiment testing that device
A runner distracted by their smartwatch
Keyboard Typing on an Unmodified Smartwatch Using Sonar
Mentor: Jiwan Kim (mail: kjwan4435@gmail.com, web: http://jiwan.kim/)
Required skills: Android programming, Python data processing
Background: Interaction with smartwatches is limited by the small size of their touch screens and occlusion issues, especially for text input. To address this limitation, many previous works explored text input on the smartwatch, but they are relying on external devices or still have occlusion issues. I want to suggest SonarType, which can sense around-movement without external sensors and occlusion issues, and with mitigating interference of nearby moving object.
Expected Outcomes: Mobile application prototyping, Data analysis, Run experiment
References: �1. Sonar sensing on the smartwatch: Kim, J., & Oakley, I. (2022, April). SonarID: Using Sonar to Identify Fingers on a Smartwatch. In Proceedings of the 2022 CHI Conference on Human Factors in Computing Systems (pp. 1-10).
2. Around gesture tracking using sonar on mobile device: Wang, W., Liu, A. X., & Sun, K. (2016, October). Device-free gesture tracking using acoustic signals. In Proceedings of the 22nd Annual International Conference on Mobile Computing and Networking (pp. 82-94).
3. Text entry for small touchscreens: Gong, J., Xu, Z., Guo, Q., Seyed, T., Chen, X. A., Bi, X., & Yang, X. D. (2018, April). Wristext: One-handed text entry on smartwatch using wrist gestures. In Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems (pp. 1-14).
WristText (CHI’18)
LLAP (MobiCom’16)
Our system
Development and Design Gaze-Pinch Input Interaction on Tablet/Desktop
Current pointing and selection method in Apple Vision Pro�Gaze(Eye) + Pinch(Hand) on Virtual Reality
What if we move this Gaze(Eye) + Pinch(Hand) on Tablet/Desktop with webcam?
Left: how to get gaze in built-in-camera in Phone
Right: Mediapipe to get hand joint point
References
[1] Huang, Qiong, Ashok Veeraraghavan, and Ashutosh Sabharwal. "Tabletgaze: dataset and analysis for unconstrained appearance-based gaze estimation in mobile tablets." Machine Vision and Applications 28.5 (2017): 445-461.
[2] Pfeuffer, Ken, and Hans Gellersen. "Gaze and touch interaction on tablets." Proceedings of the 29th Annual Symposium on User Interface Software and Technology. 2016.